基于PSO优化K均值聚类的葡萄果穗图像分割算法  

Grape Image Segmentation Algorithm Based on Particle Swarm Optimization and K-means Clustering

作　　者：张永剑 任洪娥[1] ZHANG Yongjian;REN Honge(College of Information and Computer Engineering,Northeast Forestry University,Harbin 150040,China)

机构地区：[1]东北林业大学信息与计算机工程学院,哈尔滨150040

出　　处：《智能计算机与应用》2020年第5期81-84,88,共5页Intelligent Computer and Applications

基　　金：东北林业大学国家级大学生创新创业计划训练项目(201810225177);中央高校基本科研业务费专项资金资助项目(2572017PZ10)。

摘　　要：针对复杂背景下现实葡萄果穗图像分割问题,本文提出了一种基于粒子群优化K均值聚类的分割算法,利用粒子群算法的全局搜索能力,优化初始聚类中心的选择,根据得到的聚类中心进行聚类,可以有效改善K均值聚类算法易受初始化聚类中心影响的问题。实验表明,该算法分割准确率较高,分割得到的葡萄轮廓完整,可以有效的将葡萄从复杂背景中识别出来。Aiming at the problem of segmentation of realistic grape image under complex background,a segmentation algorithm based on particle swarm optimization and K-means clustering was proposed,by using the global search capability of particle swarm algorithm to optimize the selection of initial clustering centers.Class center clustering can effectively improve the problem that K-means clustering is susceptible to initialization cluster centers.Experiments show that the proposed algorithm has high segmentation accuracy and complete grape contours,which can effectively identify grapes from complex backgrounds.

关 键 词：粒子群算法 K均值聚类算法 图像分割 

分 类 号：TP391.41[自动化与计算机技术—计算机应用技术]